Papers with Aligning large language models
Value Alignment from Unstructured Text (2024.emnlp-industry)
Copied to clipboard
Inkit Padhi, Karthikeyan Natesan Ramamurthy, Prasanna Sattigeri, Manish Nagireddy, Pierre Dognin, Kush Varshney
| Challenge: | Currently, alignment of large language models to value systems relies on the availability of supervised and preference data. |
| Approach: | They propose a systematic approach for aligning large language models to values in unstructured text data using synthetic data generation techniques. |
| Outcome: | The proposed approach shows improved performance on the Mistral-7B-Instruct model compared to other approaches, as quantified through the use of automatic metrics and win rates. |
Intention Analysis Makes LLMs A Good Jailbreak Defender (2025.coling-main)
Copied to clipboard
| Challenge: | Existing methods to align large language models with human values overlook the intrinsic nature of jailbreaks, which limits their effectiveness in complex scenarios. |
| Approach: | They propose a simple yet highly effective defense strategy, i.e., Intention Analysis (IA). They show that IA suppresses LLM’s tendency to follow jailbreak prompts, thereby enhancing safety. |
| Outcome: | The proposed strategy reduces harmfulness of LLMs and outperforms GPT-3.5 in attack success rate. |
GATEAU: Selecting Influential Samples for Long Context Alignment (2025.emnlp-main)
Copied to clipboard
Shuzheng Si, Haozhe Zhao, Gang Chen, Yunshui Li, Kangyang Luo, Chuancheng Lv, Kaikai An, Fanchao Qi, Baobao Chang, Maosong Sun
| Challenge: | Existing studies have attempted to scale up the available data volume by synthesizing long instruction-following samples, but a lack of a well-defined strategy for ensuring data quality may introduce low-quality samples and restrict the model’s performance. |
| Approach: | They propose a framework to identify influential samples enriched with long-range dependency relations that can be used to align large language models to handle instructions with extremely long contexts. |
| Outcome: | The proposed framework identifies samples with long-range dependency relations and shows that the model trained on these samples exhibits better instruction-following and long-context understanding capabilities. |
PURE: Aligning LLM via Pluggable Query Reformulation for Enhanced Helpfulness (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Large language models (LLMs) depend on vast amounts of text data sourced from the Internet for their training. |
| Approach: | They propose a new alignment paradigm that reformulates risky queries into highly relevant yet harmless ones before feeding them into LLMs. |
| Outcome: | The proposed approach eliminates the high costs of training base LLMs and achieves a promising balance of harmlessness and helpfulness. |
Direct Large Language Model Alignment Through Self-Rewarding Contrastive Prompt Distillation (2024.acl-long)
Copied to clipboard
| Challenge: | Existing methods to evaluate preference data without human annotations are difficult . et al., 2022b) is effective for aligning large language models with human expectations . |
| Approach: | They propose a method to evaluate the response preference using output probabilities under contrastive prompts. |
| Outcome: | The proposed method could surpass the RLHF method without human-annotated preference data. |
On Diversified Preferences of Large Language Model Alignment (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Large language models (LLMs) can be fine tuned with human feedback, but human preferences can be diversified due to annotators’ different tastes, which hinders the effectiveness of LLM alignment methods. |
| Approach: | They propose a calibration error metric to evaluate large language models (LLMs) and a multi-objective reward learning method to enhance the calibration performance of RMs on shared preferences. |
| Outcome: | The proposed model can be adopted as a key calibration error and MORE can achieve superior alignment performance. |
Sample Efficient Alignment Learning With Episodic Control (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing parametric methods for aligning large language models with task objectives are limited. |
| Approach: | They propose a non-parametric framework that aligns large language models with task objectives . they use a key-value memory to store associations between generated text and its corresponding values . |
| Outcome: | The proposed framework outperforms state-of-the-art baselines on harmless, helpful, and summarization tasks. |
Aligning Large Language Models with Human Preferences through Representation Engineering (2024.acl-long)
Copied to clipboard
Wenhao Liu, Xiaohua Wang, Muling Wu, Tianlong Li, Changze Lv, Zixuan Ling, Zhu JianHao, Cenyuan Zhang, Xiaoqing Zheng, Xuanjing Huang
| Challenge: | Existing methods for achieving this alignment involve employing reinforcement learning from human feedback (RLHF) Existing approaches involve using RLHF to fine-tune LLMs based on human labels . however, RLRF is susceptible to instability during fine- tuning and presents challenges in implementation. |
| Approach: | They propose to use reinforcement learning from human feedback to fine-tune large language models with human preferences to achieve precise control of model behavior. |
| Outcome: | Experiments show that RAHF can be used to capture and manipulate representations to align with a broad spectrum of human preferences or values rather than being confined to a single concept or function. |
Reward-Shifted Speculative Sampling Is An Efficient Test-Time Weak-to-Strong Aligner (2025.emnlp-main)
Copied to clipboard
| Challenge: | Recent research has focused on test-time alignment, where additional compute is allocated during inference to enhance LLM safety and reasoning capabilities. |
| Approach: | They propose a reward-shifted speculative sampling algorithm that aligns a draft model with human preferences while the target model remains unchanged. |
| Outcome: | The proposed algorithm achieves superior gold reward scores at a significantly reduced inference cost in test-time weak-to-strong alignment experiments. |
Aligning Large Language Models through Synthetic Feedback (2023.emnlp-main)
Copied to clipboard
| Challenge: | Currently, alignment learning requires significant human demonstrations and feedback from proprietary LLMs such as ChatGPT. |
| Approach: | They propose a framework that uses synthetic feedback to align large language models to human values without extensive human annotations and proprietary LLMs. |
| Outcome: | The proposed model outperforms open-source models on human-annotated demonstrations in alignment benchmarks. |
Governance in Motion: Co-evolution of Constitutions and AI models for Scalable Safety (2025.emnlp-main)
Copied to clipboard
Chenhao Huang, Ziyu Shen, Yicong Ren, Huiyuan Zheng, Jiazheng Zhang, Mingxu Chai, Ming Zhang, Shihan Dou, Fan Mo, Jie Shi, Tao Gui, Qi Zhang, Xuanjing Huang
| Challenge: | Existing approaches to align large language models with human preferences lack flexibility . static alignment preferences lack the ability to correct misaligned behaviors as they emerge . |
| Approach: | They propose a framework that enables dynamic and continuous alignment of large language models with human preferences. |
| Outcome: | The proposed framework improves safety and accuracy of a 7B model with human annotations. |
Inductive-Deductive Strategy Reuse for Multi-Turn Instructional Dialogues (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods target instruction dialogues as learning goal and fine-tune user simulators to pose instructions. |
| Approach: | They propose to use real instruction dialogues to model complex dialogue flows and pose high-quality instructions. |
| Outcome: | The proposed method generates diverse, in-depth, and insightful instructions for a given dialogue history. |
DORM: Preference Data Weights Optimization for Reward Modeling in LLM Alignment (2025.findings-emnlp)
Copied to clipboard
Rongzhi Zhang, Chenwei Zhang, Xinyang Zhang, Liang Qiu, Haoming Jiang, Yuchen Zhuang, Qingru Zhang, Hyokun Yun, Xian Li, Bing Yin, Tuo Zhao, Chao Zhang
| Challenge: | Existing approaches to align large language models with human preferences are noisy and varying in importance of preference samples. |
| Approach: | a new method enhances reward modeling by learning to dynamically weigh preference data. |
| Outcome: | a new method improves the performance of large language models with human preferences . it initializes data importance and iteratively refines them to maximize validation performance. |